Tested on RTX3090, memory optimizer off

gen_compare/README.md

@@ -2,11 +2,11 @@
 
 ## Implementation Source Code & versions
 - Diffusers (in development version): https://github.com/takuma104/diffusers/tree/e758682c00a7d23e271fd8c9fb7a48912838045c
-- Reference Impl.: https://github.com/lllyasviel/ControlNet/tree/2f77609bf6a8a2243d9faa198365fc6222c5435f
+- Reference Impl.: https://github.com/lllyasviel/ControlNet/tree/ed0439b37ce0872789e425d46381e892dbc9b407 
 
 ## Environment
 - OS: Ubuntu 22.04
-- GPU: Nvidia RTX3060 
+- GPU: Nvidia RTX3090
 - Cuda: 11.7
 - Pytorch: 1.13.1
 
@@ -39,7 +39,7 @@ All images were generated with the same prompt.
 - sampler: DDIM
 - guidance_scale: 9.0
 - num_inference_steps: 20
-- initial random latents: created on CPU using seed
+- initial random latents: created on CPU using the seed
 
 ## Results:
 

gen_compare/gen_diffusers_image.py

@@ -33,7 +33,7 @@ if __name__ == '__main__':
     model_id = f'../../control_sd15_{model_name}'
 
     pipe = StableDiffusionControlNetPipeline.from_pretrained(model_id).to("cuda")
-    pipe.enable_attention_slicing(1)
+    #pipe.enable_attention_slicing(1)
 
     image_types = {'bird', 'human', 'room', 'vermeer'}